CICSplex Plex Service Class Analysis attribute group

The CICSplex Plex Service Class Analysis attributes report on statistical information for a service class interval in your CICSplex wide environment. Use the CICSplex Plex Service Class Analysis attributes in situations to monitor performance items such as the response time goal for the service class, the number of completed transactions for an interval, average response time, and percent-of-goal information for the service class during the interval.

50 Percent of Goal Transaction Count Indicates the number of transactions whose response time was 50% (or less) of the response time goal. The value format is a positive integer, maximum four characters.

60 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 50% and less than 60% of the response time goal. The value format is a positive integer.

70 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 60% and less than 70% of the response time goal. The value format is a positive integer.

80 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 70% and less than 80% of the response time goal. The value format is a positive integer.

90 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 80% and less than 90% of the response time goal. The value format is a positive integer.

100 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 90% and less than 100% of the response time goal. The value format is a positive integer.

110 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 100% and less than 110% of the response time goal. The value format is a positive integer.

120 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 110% and less than 120% of the response time goal. The value format is a positive integer.

130 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 120% and less than 130% of the response time goal. The value format is a positive integer.

140 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 130% and less than 140% of the response time goal. The value format is a positive integer.

150 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 140% and less than 150% of the response time goal. The value format is a positive integer.

200 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 150% and less than 200% of the response time goal. The value format is a positive integer.

400 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 200% and less than 400% of the response time goal. The value format is a positive integer.

Average 1st Dispatch Delay Indicates the average time that tasks within the service class waited to be dispatched for the first time. The wait might be due to the maximum or class maximum tasks limit being reached.

Average DB2 Time Indicates the average time that tasks within a service class spent in DB2.

Average Dispatch Time Indicates the average time that tasks within the service class spent running on a CICS TCB.

Average DSA Storage occupied Indicates the average amount of user DSA storage that was occupied by the tasks within this service class for the duration of the collection interval.

Average EDSA Storage occupied Indicates the average amount of user EDSA storage that was occupied by the tasks within this service class for the duration of the collection interval.

Average Execution Time Indicates the average time from the task's first dispatch until the end of the task for the tasks within the service class.

Average Response Time Indicates the average time (accurate to three decimal places) taken to complete all tasks during the collection interval. This average is calculated by adding the response time of each task within a service class and dividing the sum by the total number of completed tasks during the collection interval. The value can be entered in the HH:MM:SS.DDD format, for example, 00:20:00.567 or as SSSSSSS.DDD, for example 1200.567. If you are using variable substitution, the value in this example would display as 1200567. The value format is a positive number, maximum four bytes.

Average Response Time MS Indicates the average time in seconds taken to complete all tasks during the collection interval.

Average time using CPU Indicates the average time that tasks within a service class spent using the CPU.

Average time using RLS CPU Indicates the average amount of CPU time that each RLS file request consumed on an MVS SRB on behalf of a given service class.

Average time waiting on Adabas Indicates the average time that tasks within a service class waited for ADABAS requests to complete.

Average Unidentifiable Delay Indicates the average time that tasks within a service class spent waiting on resources other than CPU, DB2, DL/I, File Control, MRO, Redispatch, Temporary Storage, Transient Data, RLS CPU, and ENQs.

Average Wait on Datacom Indicates the average time that tasks within a service class spent waiting for Datacom requests to complete.

Average Wait on DLI Indicates the average time that tasks within a service class spent waiting for DL/I requests to complete.

Average Wait on ENQ Delay Indicates the average time that tasks within the service class waited as a result of an enqueue.

Average Wait on FILEs Indicates the average time that tasks within a service class spent waiting for File requests to complete.

Average Wait on IC Delay Indicates the average time that tasks within a service class waited as a result of an interval control delay.

Average Wait on IDMS Indicates the average time that tasks within a service class spent waiting for IDMS requests to complete.

Average Wait on Journal Indicates the average time that tasks within the service class spent waiting for explicit journal I/O requests to complete.

Average Wait on MQ Indicates the average time that tasks within a service class waited for the MQ connection.

Average Wait on MRO Indicates the average time that tasks within the service class waited for all MRO operations to complete.

Average wait on Redispatch Indicates the average time that the tasks in the service class waited for redispatch.

Average Wait on SUPRA Indicates the average time that tasks within a service class waited for SUPRA requests to complete.

Average Wait on TD-IO Indicates the average time that tasks within a service class waited for VSAM transient data I/O.

Average Wait on TS IO Indicates the average time that tasks within a service class waited for VSAM temporary storage I/O.

CICSplex Name The name of the CICSplex that is being monitored. This name is an alphanumeric string, with a maximum of eight characters, and case-sensitive. CICSplex names are always uppercase characters.

CPU Time Std Deviation Indicates the standard deviation of the CPU time for the tasks within this collection interval.

Goal Response Time Indicates the time (accurate to three decimal places) that has been set as a goal for a service class. The value can be entered in the HH:MM:SS.DDD format, for example, 00:20:00.567 or as SSSSSSS.DDD, for example 1200.567. If you are using variable substitution, the value in this example would display as 1200567. The value format is a positive number, maximum four bytes.

Goal Type Indicates the performance objective that your site defines for a service class. This objective is expressed as an average response time (A) or a percentage of transactions meeting a specified response time goal (P). These are the values: Average and Percent.

The format for an average response time goal is a response time goal specified as an average response time is formatted in Service Level Analysis views as HH:MM:SS:TTT-hours, minutes, seconds, and thousands of seconds.

Example: An average response time goal of five and one-half seconds shows as 00:00:05:500.

The format for a percentage goal is a percentage of goal is formatted in Service Level Analysis views as HH:MM:SS:TTT-nn%. The response time goals for a service class can be overridden within a service policy. For example, 00:00:01:000-90% specifies that at least 90 percent of the transactions in the service class should finish within one second.

Greater than 400 Percent of Goal Transaction Count Indicates the number of transactions whose response time was more than 400% of the response time goal. The value format is a positive integer.

Highest Response Time Indicates the highest response time in seconds of all tasks during the collection interval.

Highest Time Using CPU Indicates the highest CPU time in seconds of all tasks during the collection interval.

Interval End Timestamp Indicates the time when the CICS Service Level Analysis collector finished accumulating service class data for the interval. The value format is CYYMMDDHHMMSSmmm, where:

C: Century (0 for 20th, 1 for 21st, and so forth)
YY: Year
MM: Month
DD: Day
HH: Hour
MM: Minutes
SS: Seconds
mmm: Milliseconds

Example: 1010521130500000 indicates the data was collected on May 21, 2001 at 1:05 p.m.

Interval Length Indicates the length of the interval in seconds at which CICS Service Level Analysis data is summarized. This interval is five minutes in length for current reports. For historical reports, this item indicates the interval set by your system administrator. The value format is a positive integer, maximum four bytes.

Interval Start Timestamp Indicates the time when the CICS Service Level Analysis collector began accumulating service class data for the interval. The value format is CYYMMDDHHMMSSmmm, where:

C: Century (0 for 20th, 1 for 21st, and so forth)
YY: Year
MM: Month
DD: Day
HH: Hour
MM: Minutes
SS: Seconds
mmm: Milliseconds

Example: 1010521130500000 indicates the data was collected on May 21, 2001 at 1:05 p.m.

Number of Abended Transactions Indicates the number of transactions within this collection interval that abnormally terminated.

Origin Node The unique identifier that describes the CICSplex. It consists of the CICSplex name and a string to identify it as a CICSplex. The value format is an alphanumeric string, maximum 32 bytes, and case-sensitive.

When a situation is true, the system compares the origin node name in the attribute to the names of managed systems assigned to the item. If the origin node name matches the name of a managed system, the status of the item changes accordingly. If the names of the origin node and the managed systems do not match, the status of the item remains unchanged.

Percent of Goal Indicates what percentage of transactions within a service class must meet the response time goal. It is assigned to a service class when the service class is created or when its goal is overridden for a service policy. The value format is a positive integer.

Percent of time Dispatched Indicates the percentage of the elapsed time that tasks within the service class spent running on a CICS TCB.

Percent of time using CPU Indicates the percentage of elapsed time that tasks within a service class spent using the CPU. The value format is a positive integer.

Percent of time using CPU for RLS Indicates the amount of CPU time that each RLS file request consumed on a z/OS SRB on behalf of a given service class. This time combined with the Percent of time using CPU give the total amount of CPU consumed by the service class for a given attribute. The value format is a positive integer.

Percent of time waiting for 1st Dispatch Indicates the percentage of elapsed time that tasks within the service class waited to be dispatched for the first time. The wait might be due to the maximum or class maximum tasks limit being reached.

Percent of time waiting on Adabas Indicates the percentage of elapsed time that tasks within a service class waited for ADABAS requests to complete. The total ADABAS wait time is expressed as a percentage of the total elapsed time.

Percent of time waiting on Datacom Indicates the percentage of elapsed time that tasks within a service class waited for DATACOM requests to complete.

Percent of time waiting on DB2 Indicates the percentage of elapsed time that tasks within a service class waited for DB2 requests to complete. The total DB2 wait time is expressed as a percentage of the total elapsed time.

Percent of time waiting on DLI Indicates the percentage of elapsed time that tasks within a service class waited for DL/I requests to complete. The total DL/I wait time is expressed as a percentage of the total elapsed time.

Percent of time waiting on ENQ Delay Indicates the percentage of elapsed time that tasks within a service class waited as a result of an enqueue.

Percent of time waiting on File Control Indicates the percentage of elapsed time that tasks within the service class waited for all file I/O requests to complete. A high percentage of elapsed time spent waiting for file control requests indicates inefficient local shared resources (LSR) specifications, too few strings, or lockout conditions. The value format is a positive integer.

Percent of time waiting on IDMS Indicates the percentage of elapsed time that tasks within a service class waited for IDMS requests to complete. The total IDMS wait time is expressed as a percentage of the total elapsed time.

Percent of time waiting on Interval Control Indicates the elapsed time that a service class waited as a result of an interval control delay. The value format is a positive integer.

Percent of time waiting on Journal Control Indicates the percentage of elapsed time that tasks in the service class waited for explicit journal I/O requests to complete. The value format is a positive integer.

Explicit journal I/O includes any journal switching, buffering, and intervention that occurs between the time the journal requests were issued and the time they completed. Updates to the file triggers journal operations, the elapsed time percentage for which is included under File Control Percentage. A high percentage of elapsed time spent waiting for journal I/O indicates disk contention or reserves, elongated journal switching times, excessive journal buffer sizes, or inappropriate journal options.

Note: This data is available only for CICS Transaction Server, V3.1 and higher.

Percent of time waiting on MQ Indicates the percentage of time that was spent waiting for the MQ connection. The value format is a positive integer.

Percent of time waiting on MRO Indicates the percentage of elapsed time that tasks in the service class waited for all multiregion operations (MROs) to complete. A high percentage of time spent waiting for MRO operations indicates problems in a connected CICS region within a CICSplex that caused the originating transaction to wait for an extended period of time. The value format is a positive integer.

Note: This data is available only for CICS Transaction Server, V3.1 and higher.

Percent of time waiting on Redispatch Indicates the percentage elapsed time that the tasks in the service class waited for redispatch. This is the aggregate of the wait times between each event completion and user-task redispatch, but it does not include the time spent waiting for first dispatch. A high percentage of elapsed time spent waiting for re dispatch indicates that you have looping transactions or that you have reached the CMXT limit. The value format is a positive integer.

Note: This data is available only in CICS Transaction Server, V3.1 and higher.

Percent of time waiting on SUPRA Indicates the percentage of elapsed time that tasks within a service class waited for SUPRA requests to complete. The total SUPRA wait time is expressed as a percentage of the total elapsed time.

Percent of time waiting on TD I/O Indicates the percentage elapsed time that the tasks in the service class waited for VSAM transient data I/O. A high percentage of elapsed time spent waiting for this type of I/O indicates inadequacies in the configuration of the transient data set DFHINTRA; this is caused by insufficient strings, bad VSAM control interval size and (CISZ) inefficient or conflicting disk allocation. The value format is a positive integer.

Note: This data is available only for CICS Transaction Server, V3.1 and higher.

Percent of time waiting on TS I/O Indicates the percentage of elapsed time that the tasks in a service class waited for VSAM temporary storage I/O. A high percentage of elapsed time spent waiting for this type of I/O indicates inadequacies in the configuration of the DFHTEMP temporary storage data set. Typical causes are insufficient strings that are not correctly defined for VSAM control interval size (CISZ) and inefficient or conflicting disk allocation. The value format is a positive integer.

Note: This data is available only for CICS Transaction Server, V3.1 and higher.

Percent of time waiting on unidentifiable Indicates the percentage of elapsed time spent by tasks in a service class waiting on resources other than: CPU, DB2, DL/I File Control, Journal, MRO, Redispatch, Temporary Storage, or Transient Data. This item includes waits on interval control, Basic Mapping Support Paging (BMS), and program loading. This percentage is calculated by subtracting the percentages listed above from 100. The value format is a positive integer.

Performance Index Indicates how well a service class is performing relative to its response time goal. The value format is a positive integer, maximum four digits.

This value is stored in units of 1/100; therefore, when you open an event as attributes, you must interpret a Performance Index value of 1234 as 12.34. When you see the Performance Index in a view, however, a decimal point is inserted; therefore you interpret a value of 1234 as 12.34.

Note: Average response-time goals: If you specify the response time goal of a service class as an average response time, the performance index measures how close the response time for the service class is to the response time goal. The performance index is calculated by dividing the actual average response time by the response time goal. When the average response time equals the response time goal, the performance index is displayed as 1.00.

For example, if the average response time for a service class is 1.5 seconds and the response time goal is two seconds, then the performance index is 0.75.

If you specify the response time goal as a Percent of Goal (also known as a percentile goal), the performance index measures whether the percentage of transactions specified in the goal meets the response time goal.

When a task completes, it is calculated in one of the following percent-of-goal buckets: 50% or less, 60%, 70%, 80%, 90%, 100%, 110%, 120%, 130%, 140%, 150%, 200%, 400%, or greater than 400%. This is based on its actual response time compared with the goal response time. The performance index is calculated by locating the percent-of-goal bucket with a cumulative total equal to or higher than the percent of goal for the service class. The cumulative total is the sum of the number of transactions for each bucket, from lowest to highest.

Example: Assume the following:

The response time goal for a service class is 80% of all transactions completing within 3 seconds.
10 transactions complete within the time interval as follows: The first three transactions average 1.5 seconds (50% of goal); the second three average 2.9 seconds (90% of goal); the third three average six seconds (200% of goal); and the last transaction completes in 99 seconds (400% of goal).
The total transaction count is 10. Eighty percent of the total transaction count is eight.

Adding from lowest to highest, the eighth transaction occurs in the 200% bucket. The performance index is, therefore, 2.00. Note that the transaction in the >400% bucket does not affect the performance index.

Performance Index Percent Indicates how well a service class is performing relative to its response time goal. The value format is a positive integer.

Response Time Std Deviation Indicates the standard deviation of the response time for the tasks within this collection interval.

Service Class Name Indicates the eight character user-defined name that identifies a service class. The value format is an alphanumeric string, maximum eight characters, and case-sensitive.

Summary Type Indicates the type of summary for the service class. The value format is an alphanumeric string, maximum four characters, and case-sensitive.

Transaction ID Indicates the four character name of the transaction for this summary. It may be null if the summary type is not transaction specific. The format is an alphanumeric string with a maximum of four characters, and is case-sensitive.

Transaction rate per minute Indicates the average number of transactions that were run in one minute of elapsed time. For example, if the statistics are collected every 10 minutes, and the total number of transactions during the last 10 minutes is 300, the average number of transactions per minute is 30. The value format is a positive integer, maximum four characters.

Transactions Total Indicates the total number of tasks in a service class that completed execution during the collection interval. The value format is a positive integer, maximum four characters.

Workload Name Indicates the eight-character work load name assigned to this service class when the service class was created. The value format is an alphanumeric string, maximum eight characters, and case-sensitive.